Digging into the roots of belowground carbon cycling following seven years of Prairie Heating and CO2 Enrichment (PHACE), Wyoming USA

Soil Biology & Biochemistry 115 (2017) 169e177

Contents lists available at ScienceDirect

Soil Biology & Biochemistry journal homepage: www.elsevier.com/locate/soilbio

Digging into the roots of belowground carbon cycling following seven years of Prairie Heating and CO2 Enrichment (PHACE), Wyoming USA Laura Nelson a, Dana M. Blumenthal b, David G. Williams a, Elise Pendall a, c, * a

Department of Botany, University of Wyoming, 1000 East University Avenue, Laramie, WY 82071, USA Rangeland Resources Research Units, USDA-ARS, Crops Research Laboratory, 1701 Centre Avenue, Fort Collins, CO 80526, USA c Hawkesbury Institute for the Environment, University of Western Sydney, Penrith, 2751 NSW, Australia b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 1 March 2017 Received in revised form 13 August 2017 Accepted 22 August 2017

Grassland soils are significant carbon (C) sinks as more than half of grassland plant biomass is belowground and roots are the main source of soil C. It is uncertain if grassland soils will continue as C sinks in the future because climate change may affect the dynamic, belowground relationships among crown and root biomass, root chemistry and morphology, and root and soil decomposition, all of which influence C sequestration potential. To better understand future belowground C cycling in semiarid grasslands we analyzed three native species (Bouteloua gracilis, Carex eleocharis, and Pascopyrum smithii) and mixedgrass community crown and root biomass, root chemistry, morphology, and decomposability, and soil organic carbon (SOC) priming following seven years of simulated climate change at the Prairie Heating and CO2 Enrichment (PHACE) experiment in Wyoming, USA. We found that individual species and the community respond uniquely to the climate change field treatments, indicating that species composition is important when analyzing climate change effects on grassland C cycling. Root biomass in the C3 sedge, C. eleocharis, increased under elevated CO2, especially when combined with warming. Decomposition rates of roots from warming plots were higher than those from ambient plots for B. gracilis and P. smithii. Across species, root decomposition rates increased with C and N concentrations. Root morphology was altered as well: B. gracilis root diameter increased under warming, and P. smithii specific root length and surface area increased under elevated CO2. P. smthii roots induced short-term, negative SOC priming across all field treatments. Together, our results indicate that grass roots may play a critical role in maintaining soil C stocks in grasslands in the future. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Elevated CO2 Warming Fine-roots Priming Carbon cycling

1. Introduction Grasslands contain more than 10% of the global carbon (C) stock and 98% of that C is in the soil (Jones and Donnelly, 2004; Heinemeyer et al., 2012). The significant amount of C in grassland soils may be attributed to the high root:shoot ratios that characterize grasslands (Mokany et al., 2006). For example, 90% of plant biomass and 67% of NPP in semiarid shortgrass steppe is concentrated belowground (LeCain et al., 2006). Additionally, temperate grasslands contain 17% of the global fine-root pool (less than 2-mm diameter) (Jackson et al., 1997). A large portion of belowground C is also stored as carbohydrates in crowns just below the soil surface

* Corresponding author. Department of Botany, University of Wyoming, 1000 East University Avenue, Laramie, WY 82071, USA. E-mail address: [email protected] (E. Pendall). http://dx.doi.org/10.1016/j.soilbio.2017.08.022 0038-0717/© 2017 Elsevier Ltd. All rights reserved.

(Milchunas and Lauenroth, 2001). The large quantity of C stored belowground in grasslands makes relationships among biomass, fine-root chemistry, morphology, and decomposition critical, because as CO2 concentrations and temperatures rise, the dynamics of these components and their impacts on grassland C cycling may change (Silver and Miya, 2001; Hui and Jackson, 2006; de Graaff et al., 2011; White et al., 2012). Biomass accumulation and plant tissue quality often influence one another, thus changes in biomass in response to climate change may be reflected in root chemistry (Craine et al., 2003). Increases in belowground biomass, root C:N, and longevity often co-occur under elevated CO2 (van Groenigen et al., 2005; Dijkstra et al., 2010; Pendall et al., 2011; Dieleman et al., 2012). This contrasts with warming treatments, where fine-root biomass decreases or does not significantly change, root C:N generally decreases, and N mineralization increases (Pendall et al., 2011; Dieleman et al., 2012; Carrillo et al., 2014). In an Australian native grassland, elevated CO2

170

L. Nelson et al. / Soil Biology & Biochemistry 115 (2017) 169e177

paired with warming increased root biomass, but root C:N responses were dependent on plant functional type and N distribution (Pendall et al., 2011). Therefore, assessing root biomass and chemistry responses to climate change by species may reveal important insights into belowground sources and sinks of carbon. Known controls over fine-root decomposition include root quality, root morphology, microbial composition, and environment (Silver and Miya, 2001; Vivanco and Austin, 2006; Birouste et al., 2012; Pilon et al., 2013; Bardgett et al., 2014; Smith et al., 2014). Root decomposition may decrease with elevated CO2 due to an increase in root C:N and/or lignin and suberin concentrations (Gorissen et al., 1995; van Groenigen et al., 2005; de Graaff et al., 2011; Pendall et al., 2011). Also under elevated CO2, N-limitation tends to favor roots with large diameters, low specific root lengths (SRL, length per unit mass), and greater root-tissue densities (RTD, mass per volume) (Dieleman et al., 2012). Roots with these characteristics tend to have long lifespans (Bardgett et al., 2014; Reich and Cornelissen, 2014; Prieto et al., 2015). In contrast to elevated CO2 effects, warming may lower C:N ratios, resulting in accelerated root decomposition (Silver and Miya, 2001; Pendall et al., 2004). The increase in N-availability with warming also appears to favor roots with small diameters, high SRLs, small RTDs, and short lifespans (Dieleman et al., 2012). However, in water-limited ecosystems, where elevated CO2 increases water content and warming induces soil desiccation (Nowak et al., 2004; Morgan et al., 2011), water-mediated effects might counter N-mediated effects (Dieleman et al., 2012; Pilon et al., 2013; Reich and Cornelissen, 2014). For example, in low moisture ecosystems, root decomposition appears to be accelerated under elevated CO2 (Allard et al., 2005; Dijkstra et al., 2008; Carrillo et al., 2014). Since both root decomposition and root morphology appear to be impacted by elevated CO2 and warming, and root morphology appears to influence root lifespan, quantifying both root-morphological traits and root decompostion rates in response to climate change may provide a key link in understanding C transfer from roots to the soil (Eissenstat et al., 2000). Fine-roots are quite labile relative to SOC and may induce a positive or negative priming effect on SOC decomposition (Mary et al., 1992; de Graaff et al., 2013). Positive priming is an increase, while negative priming is a decrease, in the amount or rate of SOC decomposition (Kuzyakov et al., 2000). Past studies show changes in the amount, quality, and morphology of fine-roots in response to elevated CO2 or warming (van Groenigen et al., 2005; Morgan et al., 2011; Pendall et al., 2011; Dieleman et al., 2012; Carrillo et al., 2014). Therefore, the effects of increased temperature and atmospheric CO2 on root dynamics and properties may impact the direction and extent of root-induced priming of SOC, ultimately affecting C storage in grassland soils. Climate change factors are expected to interactively alter species composition in native plant communities, depending on the resistance or vulnerability of individual species (Zelikova et al., 2014). After prolonged exposure to elevated CO2 and warming at the Prairie Heating and CO2 Enrichment (PHACE) experiment, grassland community composition shifted to favor subdominant, C3 species (especially the drought-tolerant sedge, Carex eleocharis), at the expense of the dominant C4 species (Bouteloua gracilis) (Zelikova et al., 2014; Mueller et al., 2016). We expected that these species shifts could be associated with changes in belowground resource availability, different root traits, and altered soil C cycling. Indeed, elevated CO2 combined with warming led to lower soil moisture and higher soil inorganic N in comparison with the control treatment (Mueller et al., 2016; Carrillo et al., 2012), larger root biomass and longer, thinner roots at the community-level (Mueller et al., 2016; Carrillo et al., 2014), and enhanced rates of SOC decomposition (Pendall et al., 2013). However, until the end of the

PHACE experiment in 2013 no species-level sampling was conducted belowground. The objective of this study was to better understand future belowground C cycling in semiarid grasslands by analyzing linkages between climate change effects on belowground biomass with fineroot chemistry, morphology, and decomposition, and SOC priming, particularly at the species-level. We predicted that 1) crown biomass responses to climate change treatments would be similar to root biomass responses for all species; 2) root chemical (C and N concentrations) and morphological responses to climate change treatments would depend on species identity; 3) root decomposition rates and priming of SOC decomposition would be related to root C and N contents and morphology (diameter, specific root length and surface area). We thus expected that intrinsic differences in root traits between native grassland species would mediate soil C cycle responses to climate change (Pendall et al., 2011; Burke et al., 2013). 2. Materials and methods 2.1. Study site This experiment was conducted at the Prairie Heating and CO2 Enrichment (PHACE) experiment located at the USDA-ARS High Plains Grassland Research Station (HPGRS), 15 km west of Cheyenne, WY (41
110 N, 104
54’ W; elevation 1930 m) (Carrillo et al., 2011). The PHACE site vegetation is classified as a northern mixed grass prairie composed of grasses and forbs, including a C3 grass species Pascopyrum smithii (Rydb.) A. Love and a C4 grass species Bouteloua gracilis (H.B.K) Lag., which together comprise 50% of the total aboveground biomass, and a sedge, Carex eleocharis L. Bailey (Dijkstra et al., 2010). Annual precipitation averaged 384 mm and average air temperature was 2.5
C in winter and 17.5
C in summer (Morgan et al., 2011; Carrillo et al., 2014). The soil was an Ascalon variant loam on the north side of the field site and an Altvan loam on the south side of the field site (Dijkstra et al., 2010; Nie et al., 2013). 2.2. Field experiment The PHACE experiment included twenty, 3.4 m diameter circular plots divided into four combinations of elevated CO2 and warming, with five replicates of each combination (Dijkstra et al., 2010; Carrillo et al., 2011). The four field treatments included ambient CO2 and ambient temperature (ct), ambient CO2 and elevated temperature (cT), elevated CO2 and ambient temperature (Ct), and elevated CO2 and elevated temperature (CT). The ambient [CO2] was approximately 400 mmol mol1 and the elevated [CO2] was 600 ± 40 mmol mol1 (Morgan et al., 2011). The elevated temperatures were 1.5
C above ambient during the day and 3.0
C above ambient at night (LeCain et al., 2015). Free air CO2 enrichment (FACE) was implemented during the growing season beginning in April 2006; temperatures were elevated year-round with infrared heaters attached 1.5 m above the ground on frames starting in March 2007 (Morgan et al., 2011; LeCain et al., 2015). The experiment was terminated in mid-July 2013, when all samples for this work were harvested. 2.3. Field sampling We collected soil samples from 0 to 5 cm depth by compositing four, 5 cm diameter cores from each of the twenty field plots. The soil was sieved with 2 mm sieves, weighed on site, and then transported back to the lab in coolers. We removed roots and aboveground litter by hand-picking the soil samples and then

L. Nelson et al. / Soil Biology & Biochemistry 115 (2017) 169e177

stored them at 2
C for nine days. Crown and root biomass of B. gracilis, C. eleocharis, and P. smithii were collected from each of the twenty treatment plots by excavating blocks of soil 10-cm deep and 25 by 25 cm area from each treatment plot and transporting the soil blocks to the laboratory in coolers. B. gracilis, C. eleocharis, and P. smithii roots were separated from each block by identifying the grass species in each block aboveground, tracing the plant through the soil block down to its crowns and roots, and then clipping the stems and roots from the crowns. We recovered on average 80% of the total crown biomass and 42% of the total root biomass in the three species studied (Supplemental Table 1). This is similar to the method used in a nearby prairie, where about 33% of roots were identifiable to species to a depth of 75 cm (LeCain et al., 2006), but we did not use hair detangler. Community-level roots and crowns, not distinguished by species, were collected in a similar way from another set of 25 by 25 cm blocks. 2.4. Biomass and root chemistry Crown and root samples were washed in deionized water for 30 min on a shaker bench and then dried at 70
C for ten days prior to weighing. The root biomass values represent minimum amounts due to breakage of roots during collection, and the 10-cm depth of harvest. Subsamples of clean, dry crowns and roots were ground in a ball mill and ashed at 550
C for 3 h, and % C ± SD 0.2, % N ± SD 0.04, and C:N were measured on a Costech 4010 Elemental Analyzer (Costech Analytical Technologies Inc., Valencia, CA, United States of America).

171

decomposition and priming by root C, we incubated soil (Soil) and soil with P. smithii roots (Soil þ Roots). For Soil, we weighed 15 g of soil into 50 ml plastic beakers, with each beaker containing soil sampled from one of the twenty field plots. For Soil þ Roots, we added 75 mg of dried 1 cm-cut P. smithii roots to 15 g of soil, combining roots and soil from the same field plot. We began the incubation by bringing water content to 60% of field capacity, compressing the mixture to simulate in situ bulk density, and then weighing and placing each beaker into a clean 500 ml glass-canning jar. All jars were sealed tightly and stored in the dark. The Soil and Soil þ Roots jars were stored at approximately 18.5
C and the root plus sand jars at approximately 20
C. Six jars containing only an empty beaker were used as controls. Headspace CO2 concentrations were determined on a LI-820 gas analyzer (Li-Cor 820, LICOR Inc., Lincoln, NE, United States of America) coupled to a Valco 6-port valve and 2-ml sampling loop. After sampling, jars were opened and vented outside with a small hand-held fan. We measured the d13C values of headspace CO2 in order to partition root and soil C contributions. For isotope sampling, 15 ml samples were injected into evacuated Exetainer vials and analyzed on a Thermo Gasbench coupled to a Thermo Delta Plus XL IRMS (ThermoScientific Bremen, Germany). Analytical precision was ±SD 0.1‰ for d13C. We sampled the Soil and Soil þ Roots jars for CO2 on days 1, 3, 5, 8, 12, 19, 29, 35, 42, 50, 63, 79, and 94, for isotopes on days 3, 12, 29, and 50, 63, and 94, and adjusted water content on days 13, 32, 57, and 75. We sampled the root plus sand jars for CO2 on days 1, 3, 6, 8, 12, 22, 30, 36, 43, and 58, for isotopes on days 3 and 12, and adjusted water content on days 13 and 34.

2.5. Root morphology 2.7. Data analysis Subsets of B. gracilis, C. eleocharis, P. smithii, and community roots from each field treatment plot were analyzed with WinRhizo Pro 2009 software (Regent Instruments Inc., Quebec, Canada). Approximately 150 mg of fresh roots were weighed, washed with DI water, and hand-placed using tweezers to minimize overlap. A clear transparency was placed on top of roots prior to scanning with an Epson Perfection 4870 scanner (Epson, Long Beach, CA, United States of America). After scanning, roots were oven-dried at 55
C and re-weighed. In WinRhizo all scanned-root images were analyzed at 800 DPI using a color analysis with three groups, two color classes each, to differentiate blue-sheet background color from roots. Prior to analyzing scanned-root images, any non-root debris was handselected on the digitized image and excluded. Following analysis, each image was inspected for error, including but not limited to, incomplete tracing of roots and root hairs and analysis of non-root components. Root parameters selected from the image analysis included total root length (mm), average root diameter (mm), total root surface area (mm2), and total root volume (mm3). From these parameters diameter (mm), RTD (mg mm3), SRL (mm mg1), and SRSA (mm2 mg1), were calculated on a dry weight basis. 2.6. Laboratory incubation setup and sampling To quantify potential decomposition, we incubated P. smithii, B. gracilis, and community roots in sand (Roots þ Sand). For each species, we added 75 mg of clean, dry roots cut into 1 cm-lengths to 15 g of muffled sand (3 h, 550
C) and 1 ml of soil slurry. The slurry consisted of 120 g of fresh soil collected from the field site at 0e5 cm depth, sieved to 2 mm, and combined with 100 ml DI water. We added soil slurry to confirm microbial activity since the roots were dried to stop autotrophic respiration; slurry controls contained 1 ml of soil slurry in 15 g of muffled sand. To quantify soil C

To assess the dynamics of the labile and resistant root C pools, a two-pool, three-parameter, exponential decay model was applied to CO2 production rates over time (after blank correction) using nonlinear curve fitting in Sigma Plot 13.0 (Eq. (1)) (Carrillo et al., 2011).

Rt ¼ Rl  eKl *t þ Rr

(1)

where Rt ¼ rate of CO2 production at time t (mg C g root d1), Rl ¼ labile root C decomposition rate, Rr ¼ resistant root C decomposition rate and Kl ¼ intrinsic decay constant of root C in the labile pool (d1). We used a two-part isotope mixing model to quantify the fraction of SOC-derived CO2 in Soil þ Roots jars for each sampling date, after correcting for blank CO2 concentrations and d13C values (Pendall and King, 2007) (Eq. (2))

fSOCCO2 ¼

d13 CSR  d13 CR d13 CS  d13 CR

! (2)

where fSOC-CO2 ¼ fraction of SOC-derived CO2, d13CSR ¼ d13C Soil þ Roots, d13CR ¼ d13C P. smithii roots plus sand, d13CS ¼ d13C Soil. We quantified priming as the difference in SOC-derived CO2 flux rate with and without roots (Eq. (3))

  RprimedSOC ¼ fSOCCO2  RSR  ðRS Þ

(3)

where RprimedSOC ¼ rate of primed C (mg C g soil1 day1), RSR ¼ decomposition rate of Soil þ Roots (mg C g soil1 day1), and RS ¼ decomposition rate of Soil (mg C g soil1 day1). We then determined cumulative SOC decomposition with and without roots over the 94-day incubation period.

172

L. Nelson et al. / Soil Biology & Biochemistry 115 (2017) 169e177

2.8. Statistics A two-factor factorial (CO2 by temperature) mixed-model, with a random block term for soil type, was used to analyze biomass, root chemistry, root morphology, and root decomposition parameters for each species and the community. Post hoc testing was performed with least square means separation with Tukey-Kramer adjustment. Pearson correlation matrices and simple linear regression were used to determine relationships between biomass, root chemistry, morphology, and decomposition traits by species and the community. Data that did not meet assumptions of equal variances following Hartley's method or normality of residuals were log, square root, or quarter power-transformed. When transformations did not equalize variances, a weighted analysis was performed (Eq. (4)).

1 Weights ¼ qffiffiffiffiffi S2ij

(4)

where S2ij were the variances of the ith level of CO2 by warming combination and jth level of root ID. Influential data points in linear regression analyses, points that met all of three requirements, studentized residuals >2, leverages >0.5, and Cook's D > 4 (n-k1)1, were analyzed in linear regression plots; points were then removed if removal resulted in change in significance and/or direction of slope. No data points were removed. A repeated measure analysis for a two-way analysis of variance (ANOVA), CO2 by temperature, with a blocking term for soil type, was used to examine the priming results. When sphericity failed, the adjusted Huynh-Feldt-Lecoutre Epsilon (H-F-L) results were used. We used a three-way ANOVA (CO2 by temperature by incubation treatment), split-plot completely randomized design to statistically compare treatment effects on cumulative SOC decomposition. Least square means separation with Tukey-Kramer adjustment was used for Post hoc analyses. All statistical analyses were performed with a of 0.05 in SAS v. 9.4 (SAS Institute Inc., Cary, NC, United States of America). Significant statistical results are presented clearly in figures and tables in order to streamline the text. 3. Results 3.1. Biomass Elevated CO2 and warming interacted to alter crown biomass of B. gracilis and C. eleocharis, but in opposite directions (Fig. 1a). While B. gracilis crowns expanded with warming alone, C. eleocharis crowns decreased with warming but only at ambient CO2 (Fig. 1a). By contrast, P. smithii crown biomass showed no significant differences (Fig. 1a). Likewise, crowns from the community harvest showed no treatment effects (P > 0.05; mean ¼ 260 ± 44.2 g m2). Root biomass was not significantly affected by warming or elevated CO2 in B. gracilis or P. smithii (Fig. 1b). In contrast, C. eleocharis root biomass increased with the main effect of elevated CO2 by 171% from ambient CO2 levels (Fig. 1b). Community level root biomass averaged 216 ± 55.2 g m2 across the treatments (P > 0.05). 3.2. Root chemistry Warming and elevated CO2 affected individual species and community root C, N, and C:N ratios in different ways (Table 1). There was a significant interaction between warming and CO2 for

Fig. 1. Average belowground biomass of B. gracilis, C. eleocharis, and P. smithii in a) crowns and b) roots from 0e10 cm depth after seven years of elevated CO2 and warming grouped into four field treatments: ct ¼ ambient CO2 and temperature, cT ¼ ambient CO2 and elevated temperature, Ct ¼ elevated CO2 and ambient temperature, and CT ¼ elevated CO2 and temperature. Values are means ± standard error; P-values are from two-way ANOVA, CO2 by temperature, and letters show means separations (Tukey-Kramer adjustment).

B. gracilis root C, with a positive effect in the combined treatment (Table 1). A warming and CO2 interaction significantly affected P. smithii root C, with an increase of 43% in the elevated CO2 treatment compared to ambient (Table 1). Both C. eleocharis and community root C were not statistically different among the four field treatments (P > 0.05). Both B. gracilis and P. smithii root N responded to the main effect of CO2 (Table 1). B. gracilis root N decreased by an average 14% whereas P. smithii root N increased by an average 8% compared to ambient CO2. Also, community root N increased under the main effect of warming by 21%. The climate change field treatments did not significantly affect root N of C. eleocharis (P > 0.05). C. eleocharis and community root C:N were altered by either elevated CO2 or warming (Table 1). C. eleocharis root C:N increased under the main effect of elevated CO2, while community root C:N decreased under the main effect of warming. There was no significant change in root C:N for B. gracilis or P. smithii across the four field treatments (P > 0.05).

3.3. Root morphology Three of the four analyzed root morphological traits, diameter, SRL, and SRSA, changed significantly with either the main effect of CO2 or the main effect of warming, and varied by species and community (Table 2). B. gracilis root diameter increased with warming. In P. smithii, SRL increased by 88%, and SRSA increased by 50%, under the main effect of elevated CO2 compared to ambient (Table 2). At the community level, root diameter decreased with elevated CO2, whereas SRL and SRSA increased with elevated CO2 (Table 2). Root morphology of C. eleocharis showed no statistical differences among field treatments (P > 0.05).

L. Nelson et al. / Soil Biology & Biochemistry 115 (2017) 169e177 Table 1 Average B. gracilis, C. eleocharis, P. smithii, and community root C concentration, N concentration, and C:N ratio at 0e10 cm depth after seven years of elevated CO2 and warming grouped into four field treatments; ct ¼ ambient CO2 and temperature, cT ¼ ambient CO2 and elevated temperature, Ct ¼ elevated CO2 and ambient temperature, and CT ¼ elevated CO2 and temperature. Values are means ± standard error, P-values are from two-way ANOVA, CO2 by temperature. Values followed by the same letters did not differ by treatment based on least squares means separation with a Tukey-Kramer adjustment. Significant differences between treatments were determined at a ¼ 0.05. Root ID

Field Tmt

Root C (%)

B. gracilis

ct cT Ct CT ct cT Ct CT ct cT Ct CT ct cT Ct CT

37.1 37.2 37.3 42.0 39.5 40.8 39.7 42.5 27.4 33.6 39.2 36.9 48.4 48.4 48.0 47.6

Root ID

2 FF CRD

Root C (%)

Root N (%)

Root C:N

B. gracilis

Block CO2 Warming CO2*Warming Block CO2 Warming CO2*Warming Block CO2 Warming CO2*Warming Block CO2 Warming CO2*Warming

0.17 0.032 0.034 0.043 0.066 0.48 0.12 0.57 0.44 <0.0001 0.15 0.0048 0.61 0.26 0.74 0.68

0.91 0.041 1 1 0.62 0.1 0.1 0.47 0.35 0.038 0.38 0.12 0.68 0.53 0.018 0.65

0.089 0.074 0.72 0.51 0.12 0.027 0.1 0.074 0.1 0.076 0.28 0.91 0.99 0.58 0.036 0.16

C. eleocharis

P. smithii

community

(1.0) (1.0) (0.3) (2.0) (1.1) (0.9) (2.2) (0.9) (0.8) (0.8) (1.4) (1.8) (0.6) (0.2) (0.3) (0.6)

Root N (%) b b b a

c b a ab

0.67 0.67 0.61 0.60 0.75 0.77 0.65 0.73 0.60 0.62 0.70 0.64 0.73 0.85 0.71 0.79

(0.02) (0.01) (0.04) (0.03) (0.04) (0.04) (0.02) (0.05) (0.04) (0.02) (0.04) (0.03) (0.01) (0.03) (0.01) (0.01)

Root C:N a a b b

b b a a b a b a

56 55 59 62 54 54 62 55 52 55 58 61 67 57 62 60

(2) (2) (3) (3) (2) (2) (2) (2) (2) (2) (3) (4) (1) (2) (4) (2)

b b a a

a b a b

ANOVA Table (Pr > F)

C. eleocharis

P. smithii

community

3.4. Root carbon decomposition Root decomposition rates fit the three-parameter decay model (Eq. (2)) well (Supplemental Fig. 1). Warming significantly increased the resistant root C decomposition rate, Rr, for B. gracilis roots by an average 24% compared to ambient temperature, and for P. smithii roots by an average 28% (Fig. 2). In contrast, community roots’ Rr marginally significantly decreased by an average 12% with elevated CO2 (Fig. 2). The other decay parameters, Rl and Kl, did not demonstrate significant climate change treatment effects. 3.5. Relationships between root chemistry, morphology and decomposition Root morphology and decomposition rates were correlated with root C and N concentrations, but mainly when both species and the community were combined (Fig. 3). Root diameter was inversely correlated with %C and %N, while SRL was positively correlated (Fig. 3aed). The decomposition rate of the resistant pool (Rr) was positively correlated with both %C and %N (Fig. 3eef), and the decomposition rate of the labile pool (Rl) followed a similar pattern (data not shown). Interestingly, root decomposition parameters were not correlated with root C:N ratios. For individual species, %C

173

in P. smithii roots was inversely correlated with diameter (P ¼ 0.003, r ¼ 0.64) and positively with SRL (P ¼ 0.004; r ¼ 0.65). We found no significant linear relationships between root morphology and chemistry in B. gracilis or C. eleocharis (P > 0.05). 3.6. Root-induced priming of SOC The d13C values of CO2 respired from Soil and Soil þ Roots varied over the course of the incubation, with a day by CO2 treatment interaction (Fig. 4a and b). The differences in d13C values between the elevated and ambient CO2 treatments were higher in the beginning of the experiment than after day 50, with and without roots. However, this treatment effect on the isotopic composition did not translate into a significant effect on priming. The average percent of SOC-derived CO2 increased over the duration of the incubation from an average 41% on day three to an average 90% by day ninety-four. (Fig. 4c). Regardless of treatment, adding P. smithii roots to soil significantly reduced cumulative SOC decomposition by 10.7% at the end of the experiment (Fig. 4d). 4. Discussion Following eight years of elevated CO2 and seven years of warming we found unique belowground responses of plant species and the community to the climate change treatments, which suggests that plant community composition is a significant factor in maintaining high soil C stocks in this semiarid grassland with climate change (Mueller et al., 2016). Increased decomposition rates of roots exposed to warming treatments (Fig. 2), associated with their higher N concentrations (Table 1; Fig. 3), may contribute to formation of soil organic matter (Pendall et al., 2004). Moreover, a destructive harvest at the end of the PHACE experiment allowed us to demonstrate for the first time that crown biomass of two native species responded strongly to climate change. The increase of C. eleocharis crowns and roots with elevated CO2, especially when combined with warming (Fig. 1), demonstrates a role for this subdominant C3 species e at the expense of the dominant C4, B. gracilis e in contributing to grassland community stability over eight years of exposure to future climate at the PHACE experiment (Zelikova et al., 2014). The negative priming by an important C3 grass, P. smithii (Fig. 4), provides evidence of plant-soil interactions in regulating ecosystem responses to climate change. This multicomponent study thus provides mechanistic insights into the regulation of stable soil C pools by native grassland species exposed to climate change. 4.1. Crown and root biomass Differences among species in belowground biomass responses to climate change appear to reflect intrinsic species adaptation strategies and alterations in soil resource availability under climate change. For example, the observed increase in B. gracilis crown biomass under warming independent of CO2 may be a reflection of B. gracilis’ resistance to drought (Weaver, 1968) and C4 photosynthetic pathway (Morgan et al., 2011), and is consistent with the expansion of its aboveground biomass with warming (Mueller et al., 2016). The positive effect of CO2 on C. eleocharis root biomass and crown biomass (with warming only) could be due to both direct photosynthetic responses, which are common in C3 species, and the increase in N with the combination of CO2 and warming, which might increase its ability to compete with more Nuse efficient C4 grasses (Mueller et al., 2016). The increase in belowground growth in C. eleocharis under climate change parallels its aboveground expansion, and suggests that this species is an important contributor to the resilience of this grassland in response

174

L. Nelson et al. / Soil Biology & Biochemistry 115 (2017) 169e177

Table 2 Average root morphology parameters (top table) and P-values (bottom table) by root type at 0e10 cm depth and following seven years of elevated CO2 and warming grouped into four field treatments; ct ¼ ambient CO2 and temperature, cT ¼ ambient CO2 and elevated temperature, Ct ¼ elevated CO2 and ambient temperature, and CT ¼ elevated CO2 and temperature. RD ¼ root diameter (mm), RTD ¼ root tissue density (mg mm3), SRL ¼ specific root length (mm mg1), and SRSA ¼ specific root surface area (mm2 mg1). Top table values are means ± standard error in parentheses and bottom table values are P-values from two-way ANOVA, CO2 by temperature. Values followed by the same letters did not differ by treatment based on least squares means separation with a Tukey-Kramer adjustment. Significant differences between treatments were determined at a ¼ 0.05. Root ID

Field Treatment

RD (mm)

RTD (mg mm3)

SRL (mm mg1)

SRSA (mm2 mg1)

B. gracilis

ct cT Ct CT ct cT Ct CT ct cT Ct CT ct cT Ct CT

0.26 0.29 0.26 0.28 0.23 0.22 0.24 0.22 0.41 0.41 0.35 0.32 0.23 0.25 0.22 0.20

0.554 0.523 0.552 0.520 0.476 0.466 0.520 0.502 0.540 0.472 0.467 0.431 0.495 0.475 0.494 0.507

35 30 34 32 56 57 44 56 16 17 25 35 50 44 55 65

28 27 28 28 39 39 32 37 19 21 26 31 36 34 37 40

C. eleocharis

P. smithii

community

(0.01) b (0.01) a (0.01) b (0.01) a (0.02) (0.01) (0.02) (0.01) (0.04) (0.04) (0.03) (0.05) (0.004)a (0.02) a (0.01) b (0.01) b

(.039) (.030) (.024) (.015) (.032) (.013) (.019) (.032) (.032) (.021) (.047) (.026) (.015) (.022) (.024) (.048)

(3) (3) (2) (2) (8) (6) (6) (5) (4) (3) (5) (9) (3) (4) (4) (7)

b b a a b b a a

(2) (2) (1) (1) (4) (2) (3) (3) (3) (2) (2) (4) (2) (1) (2) (3)

b b a a b b a a

ANOVA Table (Pr > F) Root ID

2 FF CRD

RD (mm)

RTD (mg mm3)

SRL (mm mg1)

SRSA (mm2 mg1)

B. gracilis

Block CO2 Warming CO2*Warming Block CO2 Warming CO2*Warming Block CO2 Warming CO2*Warming Block CO2 Warming CO2*Warming

0.5 0.72 0.039 0.58 0.011 0.62 0.18 0.38 0.79 0.055 0.66 0.66 0.42 0.024 0.79 0.11

0.14 0.93 0.27 0.98 0.7 0.15 0.6 0.88 0.86 0.15 0.19 0.67 0.23 0.6 0.91 0.59

0.62 0.89 0.16 0.58 0.056 0.26 0.3 0.36 0.99 0.029 0.37 0.45 0.15 0.012 0.63 0.1

0.3 0.92 0.62 0.7 0.15 0.2 0.36 0.45 0.92 0.015 0.28 0.62 0.06 0.047 0.67 0.24

C. eleocharis

P. smithii

community

to climate change (Zelikova et al., 2014).

4.2. Root chemistry

Fig. 2. Average resistant root C decomposition rate, Rr, from laboratory incubation of B. gracilis, P. smithii, and community roots collected from 0e10 cm depth following seven years of elevated CO2 and warming grouped into four field treatments; ct ¼ ambient CO2 and temperature, cT ¼ ambient CO2 and elevated temperature, Ct ¼ elevated CO2 and ambient temperature, and CT ¼ elevated CO2 and temperature. Values are means ± standard error; P-values are from two-way ANOVA, CO2 by temperature.

Soil N availability may explain many of the changes in root chemistry in response to climate change, such as increased N immobilization by microorganisms and a decrease in root N with elevated CO2, and increased N mineralization and root N under warming (Eissenstat et al., 2000; Gorissen and Cotrufo, 2000; Pendall et al., 2004; van Groenigen et al., 2005; de Graaff et al., 2006; Dijkstra et al., 2010; Pendall et al., 2011; Dieleman et al., 2012). Species-specific effects reflect different traits and strategies for nutrient acquisition. For instance, the increase in P. smithii root % C under elevated CO2 at ambient temperature may be a consequence of greater C allocation to roots in response to declining soil N availability (Hunt et al., 1996). We speculate that the relatively low root C concentrations in this species may reflect a large amount of silica, because these elements can be inversely related in grass roots due to a trade-off in defense mechanisms (Frew et al., 2016). Root %N increased in P. smithii under elevated CO2, possibly owing to access to soil N “hotspots” through construction of deeper or longer and thinner roots (Table 2) (Kemp and Williams, 1980; de Graaff et al., 2006; Dijkstra et al., 2008, 2010). Root %N decreased in B. gracilis under elevated CO2, reflecting higher nutrient use

L. Nelson et al. / Soil Biology & Biochemistry 115 (2017) 169e177

175

Fig. 3. Linear relationships between chemical and morphological properties for roots of P. smithii, B. gracilis and the community, collected at 0e10 cm depth following seven years of elevated CO2 and warming. a) root C and root diameter (RD); b) root N and RD; c) root C and specific root length (SRL); d) root N and SRL; e) root C and resistant root decomposition rate; f) root N and resistant root decomposition rate. Treatments were, ct ¼ ambient CO2 and temperature, cT ¼ ambient CO2 and elevated temperature, Ct ¼ elevated CO2 and ambient temperature, and CT ¼ elevated CO2 and temperature. P and r2-values are from simple linear regression (n ¼ 60).

efficiency by this C4 grass, and potentially increased competition with microbes. Altered root C:N under elevated CO2 and warming found in this study suggests potential changes in decomposability in the future (Birouste et al., 2012; Pendall et al., 2013). However, C:N ratios were not significant predictors of root decomposition in this study, possibly because C represents a range of compounds varying in degradability (van Groenigen et al., 2005; Freschet et al., 2012).

4.3. Root morphology Root morphological responses to climate change may provide insight into soil resource availability, as roots are the main conduits for water and nutrients to the plant (Pendall et al., 2013). The increase in diameter of B. gracilis with warming may be due to a greater abundance of suberized than non-suberized roots in our samples from warmed plots (Suseela et al., 2017). A study monitoring B. gracilis root growth found nonsuberized roots highly susceptible to turnover in dry soils, and that suberized roots have

greater root diameters than nonsuberized roots (Ares, 1976). Also, because suberin is a hydrophobic, resistant C compound that decreases water flow from roots into the soil, suberized roots may be favored in water-stressed soils (North, 1991; Meyer, 2013). The linear relationships between root %C, diameter and SRL suggests that C can be allocated for the production of longer, thinner roots, two traits that favor nutrient acquisition (Prieto et al., 2015), and is consistent with our observations of higher root N concentrations in longer, thinner roots. Both SRL and SRSA increased in P. smithii roots under the main effect of elevated CO2, possibly a consequence of soil water availability no longer constraining root exploration for soil sites with richer nutrient availability and faster N uptake (Reich and Cornelissen, 2014). At the community-level we observed a decrease in diameter and increase in SRL and SRSA under the main effect of elevated CO2, in agreement with a previous study (Carrillo et al., 2014). The production of longer and thinner roots under elevated CO2 may be a similar response as suggested for P. smithii, where an increase in soil water content enables construction of longer and thinner roots,

176

L. Nelson et al. / Soil Biology & Biochemistry 115 (2017) 169e177

over the long term depends on their decomposability. An increase in C quality, and/or increased N concentrations, under warming in both B. gracilis and P. smithii may explain the increased decomposition rate responses of resistant root C (Silver and Miya, 2001). An increase in root C:N and a change in lignin concentration were thought to explain decreased decomposition of roots grown under elevated CO2 following eight days of incubation (Gorissen et al., 1995). In contrast to our predictions, we found few relationships between root morphological parameters and decomposition rates, possibly because all roots were very fine in this semi-arid grassland. 4.5. SOC priming The stimulation of negative SOC priming by the addition of P. smithii roots to the soil may be explained by greater availability of labile-C in P. smithii roots than in SOC causing a shift in microorganisms predominantly decomposing SOC to microorganisms mainly decomposing P. smithii roots (de Graaff et al., 2013). Since we found no significant change in root-induced negative priming of SOC among climate change treatments, short-term negative priming of SOC may persist to the same observed degree in the future, which may lead to SOC sequestration. However, extending our results to the field is complex because previous work at our site revealed that subdominant species increase in productivity, while dominant species (including P. smithii) decrease in productivity under climate change (Zelikova et al., 2014). Thus, a change in community composition may alter P. smithii root-induced priming effects. Also, we quantified priming during a three-month period under optimal lab conditions. Therefore, long-term storage of SOC due to root-induced priming may not occur in the field, where other factors, such as precipitation, temperature, competition, and microbial dynamics fluctuate. Acknowledgments

Fig. 4. Changes in d13C of CO2 and priming of soil organic carbon (SOC) during the 94day laboratory incubation decomposition. a) d13C of CO2 generated from Soil; b) d13C of CO2 generated from Soil þ Roots; c) percentage of SOC-derived CO2; and d) rate of SOC priming (mgC gsoil1 day1). Soil was collected from 0e5 cm and P. smithii roots from 0e10 cm depth from four different field treatments applied for seven years; ambient CO2 and temperature, ct, ambient CO2 and warmed, cT, elevated CO2 and ambient temperature, Ct, and combined elevated CO2 and warmed, CT. Values are means ± standard error and P-values are from three-way ANOVA, CO2 by temperature by time (incubation day).

broadening the search for available N (Dijkstra et al., 2010; Reich and Cornelissen, 2014). Alternatively, the longer and thinner roots of the community may be driven by the increase in C. eleocharis root biomass under elevated CO2, as C. eleocharis has comparatively longer and thinner roots than either B. gracilis or P. smithii (Table 2). However, the overall different morphological responses between community-level and species-level roots to climate change suggest that an analysis of root traits at the community-level may mask species-level responses. Future studies in similar perennial grasslands would benefit from more comprehensive recovery and analysis of roots from all species present.

4.4. Root decomposition In soils, resistant C pool dynamics strongly influence soil C sequestration (Reid et al., 2012), and transfer of root C to SOC pools

We thank Y. Carrillo for insightful discussions; D. LeCain, L. Macdonald, M. Nie, S. Raut, D. Smith, S. Tamang, T.J. Zelikova andY. Carrillo for their assistance in the field and lab; and C. Macdonald and D. Wolfe at the Stable Isotope Facility at the University of Wyoming for help with sampling and analysis. Research for this project was funded by the U.S. Department of Agriculture, the USDA-CSREES Soil Processes Program (#2008-35107-18655), the U.S. Department of Energy Office of Science (BER) through the Terrestrial Ecosystem Science program (#DE-SC0006973), the Western Regional Center of the National Institute for Climatic Change Research, and the National Science Foundation (DEB#1021559). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.soilbio.2017.08.022. References Allard, V., Newton, P.C.D., Lieffering, M., Soussana, J.F., Carran, R.A., Matthew, C., 2005. Increased quantity and quality of coarse soil organic matter fraction at elevated CO2 in a grazed grassland are a consequence of enhanced root growth rate and turnover. Plant and Soil 276, 49e60. Ares, J., 1976. Dynamics of the root system of blue grama. Journal of Range Management 29, 208e213. Bardgett, R.D., Mommer, L., De Vries, F.T., 2014. Going underground: root traits as drivers of ecosystem processes. Trends in Ecology & Evolution 29, 692e699. Birouste, M., Kazakou, E., Blanchard, A., Roumet, C., 2012. Plant traits and decomposition: are the relationships for roots comparable to those for leaves? Annals of Botany 109, 463e472. Burke, I.C., Bontti, E.E., Barrett, J.E., Lowe, P.N., Lauenroth, W.K., Riggle, R., 2013. Impact of labile and recalcitrant carbon treatments on available nitrogen and

L. Nelson et al. / Soil Biology & Biochemistry 115 (2017) 169e177 plant communities in a semiarid ecosystem. Ecological Applications 23, 537e545. Carrillo, Y., Dijkstra, F.A., LeCain, D., Morgan, J.A., Blumenthal, D., Waldron, S., Pendall, E., 2014. Disentangling root responses to climate change in a semiarid grassland. Oecologia 175, 699e711. Carrillo, Y., Pendall, E., Dijkstra, F.A., Morgan, J.A., Newcomb, J.M., 2011. Response of soil organic matter pools to elevated CO2 and warming in a semi-arid grassland. Plant and Soil 347, 339e350. Carrillo, Y., Dijkstra, F.A., Pendall, E., Morgan, J.A., Blumenthal, D.M., 2012. Controls over soil nitrogen pools in a semiarid grassland under elevated CO2 and warming. Ecosystems 15, 761e774. Craine, J.M., Wedin, D.A., Chapin, F.S.I., Reich, P.B., 2003. The dependence of root system properties on root system biomass of 10 North American grassland species. Plant and Soil 250, 39e47. de Graaff, M.-A., Schadt, C.W., Rula, K., Six, J., Schweitzer, J.A., Classen, A.T., 2011. Elevated CO2 and plant species diversity interact to slow root decomposition. Soil Biology and Biochemistry 43, 2347e2354. de Graaff, M.-A., van Groenigen, K.-J., Six, J., Hungate, B., van Kessel, C., 2006. Interactions between plant growth and soil nutrient cycling under elevated CO2: a meta-analysis. Global Change Biology 12, 2077e2091. de Graaff, M.A., Six, J., Jastrow, J.D., Schadt, C.W., Wullschleger, S.D., 2013. Variation in root architecture among switchgrass cultivars impacts root decomposition rates. Soil Biology & Biochemistry 58, 198e206. Dieleman, W.I., Vicca, S., Dijkstra, F.A., Hagedorn, F., Hovenden, M.J., Larsen, K.S., Morgan, J.A., Volder, A., Beier, C., Dukes, J.S., King, J., Leuzinger, S., Linder, S., Luo, Y., Oren, R., De Angelis, P., Tingey, D., Hoosbeek, M.R., Janssens, I.A., 2012. Simple additive effects are rare: a quantitative review of plant biomass and soil process responses to combined manipulations of CO2 and temperature. Global Change Biology 18, 2681e2693. Dijkstra, F.A., Blumenthal, D., Morgan, J.A., 2010. Contrasting effects of elevated CO2 and warming on nitrogen cycling in a semiarid grassland. New Phytologist 187, 426e437. Dijkstra, F.A., Pendall, E., Mosier, A.R., King, J.Y., Milchunas, D.G., Morgan, J.A., 2008. Long-term enhancement of N availability and plant growth under elevated CO2in a semi-arid grassland. Functional Ecology 22, 975e982. Eissenstat, D.M., Wells, C.E., Yanai, R.D., Whitbeck, J.L., 2000. Building roots in a changing environment: implications for root longevity. New Phytologist 147, 33e42. Freschet, G.T., Aerts, R., Cornelissen, J.H.C., 2012. A plant economics spectrum of litter decomposability. Functional Ecology 26, 56e65. Frew, A., Powell, J.R., Sallam, N., Allsopp, P.G., Johnson, S.N., 2016. Trade-offs between silicon and phenolic defenses may explain enhanced performance of root herbivores on phenolic-rich plants. Journal of Chemical Ecology 42, 768e771. Gorissen, A., Cotrufo, M.F., 2000. Decomposition of leaf and root tissue of three perennial grass species grown at two levels of atmospheric CO2 and N supply. Plant and Soil 224, 75e84. Gorissen, A., van Ginkel, J.H., Keurentjes, J.J.B., Van Veen, J.A., 1995. Grass root decomposition is retarded when by Gorissen.pdf. Soil Biology & Biochemistry 27, 117e120. , B., Gelsomino, A., 2012. Partitioning of Heinemeyer, A., Tortorella, D., Petrovi cova soil CO2 flux components in a temperate grassland ecosystem. European Journal of Soil Science 63, 249e260. Hui, D., Jackson, R.B., 2006. Geographical and interannual variability in biomass partitioning in grassland ecosystems: a synthesis of field data. New Phytologist 169, 85e93. Hunt, H.W., Elliott, E.T., Detling, J.K., Morgan, J.A., Chen, D.X., 1996. Responses of a C3 and a C4 perennial grass to elevated CO2 and temperature under different water regimes. Global Change Biology 2, 35e47. Jackson, R.B., Mooney, H.A., Schulze, E.-D., 1997. A global budget for fine root biomass, surface area, and nutrient contents. Proceedings of the National Academy of Sciences of the United States of America 94, 7362e7366. Jones, M.B., Donnelly, A., 2004. Carbon sequestration in temperate grassland ecosystems and the influence of management, climate and elevated CO2. New Phytologist 164, 423e439. Kemp, P.R., Williams, G.J., 1980. A physiological basis for niche separation between Agropyron smithii (C00 3) and Bouteloua gracilis (C00 4). Ecology 61, 846e858. Kuzyakov, Y., Friedel, J.K., Stahr, K., 2000. Review of mechanisms and quantification of priming effects. Soil Biology & Biochemistry 32, 1485e1498. LeCain, D., Smith, D., Morgan, J., Kimball, B.A., Pendall, E., Miglietta, F., 2015. Microclimatic performance of a free-air warming and CO2 enrichment experiment in windy Wyoming, USA. PLoS One 10, e0116834. LeCain, D.R., Morgan, J.A., Milchunas, D.G., Mosier, A.R., Nelson, J.A., Smith, D.P., 2006. Root biomass of individual species, and root size characteristics after five years of CO2 enrichment on native shortgrass steppe. Plant and Soil 279, 219e228. Mary, B., Mariotti, A., Morel, J.L., 1992. Use of 13C variations at natural abundance

177

for studying the biodegradation of root mucilage, roots and glucose in soil. Soil Biology and Biochemistry 24, 1065e1072. Meyer, C.J.a.P., C.A., 2013. Structure and function of three suberized cell layers; Epidermis, Exodermis, and Endodermis. In: Eshel, A.a.B., T. (Ed.), Plant Roots: the Hidden Half, fourth ed. Taylor and Francis Group, Boca Raton, FL. 5e1-5-20. Milchunas, D.G., Lauenroth, W.K., 2001. Belowground primary production by carbon isotope decay and long-term root biomass dynamics. Ecosystems 4, 139e150. Mokany, K., Raison, R.J., Prokushkin, A.S., 2006. Critical analysis of root : shoot ratios in terrestrial biomes. Global Change Biology 12, 84e96. Morgan, J.A., LeCain, D.R., Pendall, E., Blumenthal, D.M., Kimball, B.A., Carrillo, Y., Williams, D.G., Heisler-White, J., Dijkstra, F.A., West, M., 2011. C4 grasses prosper as carbon dioxide eliminates desiccation in warmed semi-arid grassland. Nature 476, 202e205. Mueller, K.E., Blumenthal, D.M., Pendall, E., Carrillo, Y., Dijkstra, F.A., Williams, D.G., Follett, R.F., Morgan, J.A., 2016. Impacts of warming and elevated CO2 on a semiarid grassland are non-additive, shift with precipitation, and reverse over time. Ecology Letters 19, 956e966. Nie, M., Pendall, E., Bell, C., Gasch, C.K., 2013. Positive climate feedbacks of soil microbial communities in a semi-arid grassland. Ecology Letters 16, 234e241. North, G.B.a.N.P.S., 1991. Changes in hydraulic conductivity and anatomy caused by drying and rewetting roots of agave deserti (Agavaceae). American Journal of Botany 78, 906e915. Nowak, R.S., Ellsworth, D.S., Smith, S.D., 2004. Functional responses of plants to elevated atmospheric CO2- do photosynthetic and productivity data from FACE experiments support early predictions? New Phytologist 162, 253e280. Pendall, E., Bridgham, S., Hanson, P.J., Hungate, B., Kicklighter, D.W., Johnson, D.W., Law, B.E., Luo, Y., Megonigal, J.P., Olsrud, M., Ryan, M.G., Wan, S., 2004. Belowground process responses to elevated CO2 and temperature: a discussion of observations, measurement methods, and models. New Phytologist 162, 311e322. Pendall, E., Heisler-White, J.L., Williams, D.G., Dijkstra, F.A., Carrillo, Y., Morgan, J.A., LeCain, D.R., 2013. Warming reduces carbon losses from grassland exposed to elevated atmospheric carbon dioxide. Plos One 8. Pendall, E., King, Y.J., 2007. Soil organic matter dynamics in grassland soils under elevated CO2: insights from long-term incubations and stable isotopes. Soil Biology & Biochemistry 39, 2628e2639. Pendall, E., Osanai, Y.U.I., Williams, A.L., Hovenden, M.J., 2011. Soil carbon storage under simulated climate change is mediated by plant functional type. Global Change Biology 17, 505e514. Pilon, R., Picon-Cochard, C., Bloor, J.M.G., Revaillot, S., Kuhn, E., Falcimagne, R., Balandier, P., Soussana, J.F., 2013. Grassland root demography responses to multiple climate change drivers depend on root morphology. Plant and Soil 364, 395e408. Prieto, I., Roumet, C., Cardinael, R., Dupraz, C., Jourdan, C., Kim, J.H., Maeght, J.L., Mao, Z., Pierret, A., Portillo, N., Roupsard, O., Thammahacksa, C., Stokes, A., Cahill, J., 2015. Root functional parameters along a land-use gradient: evidence of a community-level economics spectrum. Journal of Ecology 103, 361e373. Reich, P.B., Cornelissen, H., 2014. The world-wide ‘fast-slow’ plant economics spectrum: a traits manifesto. Journal of Ecology 102, 275e301. Reid, J.P., Adair, E.C., Hobbie, S.E., Reich, P.B., 2012. Biodiversity, nitrogen deposition, and CO2 affect grassland soil carbon cycling but not storage. Ecosystems 15, 580e590. Silver, W.L., Miya, R.K., 2001. Global patterns in root decomposition: comparisons of climate and litter quality effects. Oecologia 129, 407e419. Smith, S.W., Woodin, S.J., Pakeman, R.J., Johnson, D., van der Wal, R., 2014. Root traits predict decomposition across a landscape-scale grazing experiment. New Phytologist 203, 851e862. Suseela, V., Tharayil, N., Pendall, E., Rao, A., 2017. Warming and elevated CO2 alter the suberin chemistry in roots of photosynthetically divergent grass species. Annals of Botany in Revision. van Groenigen, K.-J., Gorissen, A., Six, J., Harris, D., Kuikman, P.J., van Groenigen, J.W., van Kessel, C., 2005. Decomposition of 14C-labeled roots in a pasture soil exposed to 10 years of elevated CO2. Soil Biology and Biochemistry 37, 497e506. Vivanco, L., Austin, A.T., 2006. Intrinsic effects of species on leaf litter and root decomposition: a comparison of temperate grasses from North and South America. Oecologia 150, 97e107. Weaver, J.E., 1968. Prairie Plants and Their Environment: a Fifty-year Study in the Midwest. University of Nebraska Press, Lincoln, NE. White, S.R., Carlyle, C.N., Fraser, L.H., Cahill Jr., J.F., 2012. Climate change experiments in temperate grasslands: synthesis and future directions. Biology Letters 8, 484e487. Zelikova, T.J., Blumenthal, D.M., Williams, D.G., Souza, L., LeCain, D.R., Morgan, J., Pendall, E., 2014. Long-term exposure to elevated CO2 enhances plant community stability by suppressing dominant plant species in a mixed-grass prairie. PNAS. http://dx.doi.org/10.1073/pnas.1414659111.